What is Digg?124 Comments
- MCA2142, on 06/24/2008, -0/+71"Whatever Happened to Artificial Intelligence? "
Probably being used to crack CAPTCHAS - MakiMaki, on 06/23/2008, -1/+65The entire article on one-page can be found here: http://tinyurl.com/3qcoay
Tried submitting it instead but Digg said the 'URL was too long'. - swordedge, on 06/24/2008, -3/+42the AI woke up, looked at humanity, then committed suicide
- inactive, on 06/24/2008, -1/+29and then used to send Viagra spam to Yahoo Mail users
- reuscel, on 06/24/2008, -2/+25"Whatever Happened to Artificial Intelligence?"
The same question could be asked about natural intelligence. - inactive, on 06/24/2008, -1/+20We are the A.I. Dun dun dun.
- NovaPrime9, on 06/24/2008, -0/+19Screw that. I want them to work on Artificial Stupidity. Then machines can truly pass themselves off as people.
- Krumm, on 06/24/2008, -0/+14It could have been a good movie, but Jude Law hammed it up, and the ending was written by the crackhead down the street that talks to drain covers...
- houndeyex, on 06/24/2008, -1/+15Glad he narrowed it down to only 50 years. I'll either be crossing my fingers until I'm 35 or until I'm 85.
- subliminalurge, on 06/24/2008, -1/+14If only it was just Yahoo Mail users.......
- Chebsi, on 06/24/2008, -0/+13Booo.
- alecks, on 06/24/2008, -1/+13Did u try submitting the tinyurl?
- inactive, on 06/23/2008, -3/+15I talked to some researchers about it, and they seem more keen about actually understanding the computational processes and abilities of the human brain, before working too much on the actual applications.
Give it time. Kurzweil predicted a Turing-test-passing intelligence would come between 2020 and 2070 :) - Shaman760, on 06/24/2008, -0/+9Natural stupidity won out. That's what happened.
- GalacticRerun, on 06/23/2008, -2/+11Didn't they give up after it's only application - Rise of the Robots?
- inajeep, on 06/24/2008, -0/+8Someone might want to tweak that AI for loan approvals.
- LocalScope, on 06/24/2008, -0/+8So we won?
- santaliqueur, on 06/24/2008, -1/+8I'm surprised you know what the I stands for.
- merlin5, on 06/24/2008, -0/+7Happens to be the main interest of my research. For the most part Bayesian logic dominates most of the thinking lately. Its only a matter of time before Turings test is met. I beleive that "Artificial Intelligence" will be something far more complex than simply fooling humans into thinking that computers are real people.
My brother fell in love with a chat-bot ten years ago, broke his heart when I got her/it to send me the same pictures. - slapshot24, on 06/24/2008, -0/+6We figured out that AI is REALLY difficult.
Back in the 1960s, a computer science adviser at MIT thought that human vision (a pretty narrow problem for AI) could be solved in about three months of work. That led to the infamous "summer vision project," where one poor student was assigned the job of making an effective computer vision system over one summer. Finally, 40 years later, we're starting to get close to the goals of that project. - Rocco03, on 06/24/2008, -0/+5Except for rapidshare's kittens CAPTCHAS.
That's what turned Skynet into a genocide bitch. - stinkymountain, on 06/23/2008, -1/+6
awesome story...love the slideshow - VaporBro, on 06/24/2008, -0/+5I would say that you should treat this as a turning point in your life: Realize that the amount sunshine you are lacking my be interrupting your superior brain functions...or you are 8 years old. Either way; please stop stealing my air. I need it.
- truthhurts28, on 06/24/2008, -1/+6Whatever happened to regular intelligence?
- lateralus, on 06/24/2008, -0/+4Poor artificial intelligence on Digg's part.
- DuffyDirect, on 06/24/2008, -1/+5learning is a pretty basic trait period. Even ants and dogs can learn. Until a computer can, it isn't very intelligent, it's just another tool.
- Proctor, on 06/24/2008, -0/+4I met Tim Berners-Lee a few weeks ago at RPI during the Tetherless World Constellation. It was about the history of the Internet (being the creator of WWW) and Web 3.0. The entire debate section was about Artificial Intelligence in Web 3.0. Nothing happened to it, and you'll see it more and more being used in future web.
- Gudeldar, on 06/24/2008, -0/+4Movie fueled paranoia, holding science back since the 1920s!
- inajeep, on 06/24/2008, -0/+4I'll just pop over to the local radio shack and order a few trillion transistors. Maybe they'll give me a discount on all the solder and flux I need.
- houndeyex, on 06/24/2008, -0/+4...and then help me get rich by simply giving them my bank account number to store some money in the US!
- CrazedLeper, on 06/24/2008, -0/+3Perhaps, but I'm more concerned with the loss of natural intelligence.
- hartley, on 06/24/2008, -0/+3I'm pretty sure AI is still around *at least in the way originally planned*. We have had 2 examples just yesterday here on digg. The robot that played table hockey, and the full living room cleaning "ReadyBot". Both of those over 50 years ago would definitely be considered intelligent machines.
- fas2, on 06/24/2008, -2/+5That is not AI. A basic AI trait is the ability to learn. Something that is done the same way every time is by no means intelligent, even if that action is very complex.
- sarchosis, on 06/24/2008, -1/+4I don't think we really want self-aware machines. They'll be so confused about their identities that they won't get any work done.
- inactive, on 06/24/2008, -0/+3then again, he has been eerily accurate in the past
- bincoder, on 06/24/2008, -2/+5Make a matrix of a few trillion neurons using transistors, tubes, or biological cells.
Train it as a child, voila! Don't analyse it or attempt to figure out what each circuit is doing, its beyond comprehension. That would work very well, just a bit pricey.
Or just have a kid and raise it, same thing, less expensive. - id000001, on 06/24/2008, -0/+3only if you considers spam mails have nothing to do with world domination.
- mmmunaf, on 06/24/2008, -0/+3AI in the past few decades has focused on finding mathematical shortcuts to simulating human BEHAVIOR, not intelligence. Yes, machines will one day be able to pass the Turing Test. But they won't exhibit any signs of actual intelligence. Passing the Turing Test alone means nothing. Researchers need to focus on modeling ACTUAL cognitive processes, not trying to make machines exhibit intelligent BEHAVIOR.
- CasinoJack, on 06/24/2008, -0/+3I don't know much about neuroscience, but I'm pretty sure that it isn't the neurons that create intelligence, but the synapses which connect them (humans have between 0.1 and 100 trillion neurons but 1-10 quadrillion synapses).
Obviously the ultimate goal would be to map the exact structure of a human brain onto a computer chip, but this is a fair few years off yet (unless Kurzweill is right, in which case we'll be able to plug into The Matrix around 2020-2030). - Rikkochet, on 06/24/2008, -3/+6Kurzweil also made a great argument that AI is alive and well - as soon as an "experimental artificial intelligence" technology becomes reliable, it gets integrated into an existing system and isn't considered AI anymore.
Your car reacts to your braking pattern to apply an intelligent amount of braking force on individual tires. Your microwave can automatically heat your food to a servable temperature by detecting steam levels rising from the heating food.
This is all very powerful AI, but it's become expected and is now just "part of how something works". - BoomShake007, on 06/24/2008, -3/+6Possibility of a Cylon revolution too great to pursue AI further
- inactive, on 06/24/2008, -3/+6The very definition of intelligence is lacking and the so called "Turing test" is not a true indication of machine intelligence in any case. The best we will probably ever achieve without a full blown duplication of a natural brain is "simulated intelligence". Then we can pat ourselves on the backs and pretend we invented artificial intelligence........
- phreak79, on 06/24/2008, -0/+2I dare say it's still progressing but in less public ways than a generation of science fiction predicted. There was an interesting piece in the Economist this week about the robotics industry, and those are becoming smarter and more useful in a whole range of contexts.
- ligyron, on 06/24/2008, -1/+3Train "it"? How do expect a human to train "it" if "it" shares none of our five senses?
- MacEnvy, on 06/24/2008, -0/+2I wouldn't say it's "beyond comprehension". I would just say it's beyond our current understanding. I have no doubt we'll eventually be able to figure it out.
- bobangitanov2, on 06/24/2008, -0/+2I'm so glad you talked to researchers.
- jackusage, on 06/24/2008, -0/+2AI since then has been distributed computing, having multiple AI working as a team and instead of work simultaneously giving individual AI freewill and the option to under-perform or over-perform and fail their task (ex. robots playing soccer). The main problem with distributed computing is the trust element, getting individual AI to trust each other with completing parts of a specific task, all with as little human interaction as possible.
- pw378, on 06/24/2008, -0/+2You still require a boot-loader. Even a newborn baby has a pre-installed BIOS, micro-Kernel and I/O drivers. Right out of the womb, a baby knows how to feed from a breast, cry to get attention, utilize 5 senses to obtain information, and even create smelly output.
Just connecting a few trillion transitors will not create intelligence. - daveisfera, on 06/24/2008, -0/+2That 'hockey stick growth' analogy that the article uses is more commonly referred to as 'exponential growth', but I guess that would just be too big of a word for his intended audience.
-
Show 51 - 100 of 125 discussions
Check out the new & improved 
© Digg Inc. 2009
— Content created and posted by Digg users is